1. Field of the Invention
The present invention relates generally to the field of electron beam devices and more particularly to the reduction of the charging effect on a non-conductive sample and reduction of carbon deposition on samples.
2. Description of the Related Art
This invention is primarily directed toward devices which emit an electron beam over a target. Such devices are referred to herein as electron beam devices. An important example of such a device is a scanning electron microscope, which shall be referred to herein as a SEM. Electron lithography systems and electron microprobes are among other devices employing electron beams.
In a SEM, an electron beam is swept over a target and electrons that xe2x80x9cbouncexe2x80x9d off the target are collected to generate a signal representing the topographical features of the target. Both the electron beam and the sample are inside a vacuum chamber. The electrons may be back-scattered primary electrons (those electrons which are reflected back along the path they traveled during emission) or secondary electrons (electrons which are generated upon target impact). The resolution of the SEM depends in part upon the narrowness of the electron beam and the accuracy with which the beam position is controlled during the scanning operation.
When the sample to be imaged has a non-conducting surface, there is a build up of a negative charge on the sample due to the electron beam. The negative charge diverts the beam (causing a decreased beam positioning accuracy) and causes the beam to widen. Both of these effects reduce the accuracy of the SEM. An analogous problem exists in electron beam lithography processes. Such effects are referred to herein as xe2x80x9ccharging effects.xe2x80x9d
One known solution to this problem is to coat the surface of the sample with a thin conductive layer. For example, use of such a coating is discussed in U.S. Pat. No. 4,249,077, entitled xe2x80x9cION CHARGE NEUTRALIZATION FOR ELECTRON BEAM DEVICES.xe2x80x9d A second known solution involves emitting positive ions into the vacuum chamber in an attempt to neutralize the charge on the sample. This technique is also described in U.S. Pat. No. 4,249,077. Another attempted solution to this problem is to keep the intensity of the electron beam low and to keep the scan time short in order to minimize the charge that is built up on the sample. The aforementioned solutions have not proven satisfactory. They involve increased cost, complexity and/or time or are simply not sufficiently effective.
The charging effect has become a particularly serious problem in the semiconductor manufacturing field. Quartz, a non-conducting material, is often used as a substrate for masks used in photo and x-ray lithography processes. Optical microscopes have traditionally been used to review masks for defects (such as divot or bump defects on a phase shift mask and opaque and clear defects on a Cr mask) and to measure critical dimensions of masks. However, with the mask feature size now reaching the submicron level (i.e. less than 0.5 microns), optical measurements have proven inadequate. Therefore, the semiconductor industry has turned to SEMs as an alternative to optical microscopes. This reliance on SEMs for quartz mask inspection has served to highlight the deficiencies of the aforementioned techniques for reducing the charging effect on non-conductive samples.
A second, well known phenomena associated with electron beam use in general and SEMs in particular is carbon deposition, including carbon film and carbon halos, in the area near the electron beam (the image window area in a SEM). Carbon deposits may form on conductive as well as non-conductive samples. The carbon deposit is formed by electron beam bombardment of residual organic molecules inside the vacuum chamber from sources such as improperly handled samples, vacuum pump grease, etc. Although the use of proper sample handling procedures and advances in vacuum pump technology can help to reduce the amount of organic residue, to date the total elimination of organic residue is not yet possible and would most likely be prohibitively expensive even if it were. Carbon deposition adversely affects mask quality, and may cause shorts or may cause the rejection of the mask.
What is needed is a simple, inexpensive way to reduce the charging effect of electron beam devices on non-conducting samples and to reduce carbon deposition on samples of all types.
The present invention provides a method for reducing the charging effect of electron beam devices on non-conducting samples that involves introducing a water containing gas on the sample surface while the electron beam is directed on the sample. Because the water containing gas is conductive, the charge is dissipated. The water containing gas may be introduced with an adjustable nozzle and the pressure at which the gas is provided may be adjusted to provide an amount of water containing gas sufficient to dissipate the charging effect produced by the electron beam. In a preferred embodiment, the water containing gas is water vapor and the water vapor is introduced continuously. This technique is especially useful for quartz samples such as quartz photomasks because water vapor exhibits good adhesion to quartz surfaces, which helps to distribute and dissipate the charge quickly.
A second advantage of introducing water containing gas is that carbon deposits can be minimized. When a water containing gas is present, carbon monoxide or carbon dioxide is formed as the residual organic molecules react with the water molecules contained in the gas. This benefit may be realized with both conductive and non-conductive samples.
The use of water vapor to increase the material removal rates in chemically enhanced focused ion beam micro-machining is described in xe2x80x9cH2O Enhanced Focused Ion Beam Micro-machining,xe2x80x9d Stark et al., J. Vac. Sci. Technol. B 13(6), November/December 1995, p. 2565. Furthermore, the use of a water containing gas to enhance the removal rate of a carbon halo formed on a photomask during a clear defect repair with a focused ion beam is described in my co-pending application entitled xe2x80x9cMethod for Removing the Carbon Halo Caused by FIB Clear Defect Repair of a Photomask,xe2x80x9d Ser. No. 09/190057, filed Nov. 12, 1998. However, the applicant is not aware of any information that teaches use of a water containing gas to reduce the charging effect on non-conductive samples or to retard the formation of carbon deposits during exposure to electron beams or focused ion beams.